The role of product and issue involvement on the relationship between deceptive product attributes and the purchase intention of health food.

(1)

The role of product and issue involvement on the

relationship between deceptive product attributes and the

purchase intention of health food.

January 2014 Thesis final version

Author: Constantijn van Limburg Stirum (10506829) Msc. in Business Studies – Marketing

University of Amsterdam, Faculty of Economics and Business Supervisor: Frederik Situmeang

(2)

Abstract

Deceptive product attributes are used by fabricants to influence the purchase intention of consumers. Deceptiveness makes the product attributes appear to be meaningful, while in fact the product attributes are irrelevant to this benefit. Overall, product attributes are an effective way to differentiate products from competitor’s products. In a society increasingly focussed on the consumption of health food the presence of deceptive attributes also increase. Products can cause an impression of being healthy when in fact they are not. In this research area, there is still a lot unknown for both consumers and fabricants. The lack of knowledge and the availability of new product-categories leaves room for marketing purposes and especially the use of deceptive product attributes. Consumers can be misled by names, pictures, logo’s and slogans on a product’s package. The purpose of this paper is to contribute to marketing practices and increase consumer awareness towards healthy products. A research question and a set of hypotheses are based on theories from existing literature. First of all, the effect of deceptive product attributes on purchase intention is researched. Subsequently, the effects of both product involvement and issue involvement on this relationship between deceptive product attributes and purchase intention is investigated. This paper provides a theoretical framework that can be used to better understand the interaction between deceptive attributes, purchase intention and involvement and to place this in a context regarding healthy products. Results of this study suggest that deceptive product attributes have a positive effect on purchase intention. This effect seems to exist, despite the awareness of the deceptive nature of the product attributes and a lower level of healthiness of the product. These results are notable because respondents do claim to be aware of deceptive product attributes and to have a great preference for healthy products. Next to these results, research on product and issue involvement also provide noteworthy results stating that the level of involvement does not have a significant effect on the relationship between deceptive product attributes and purchase intention.

Keywords: deceptive product attributes, purchase intention, product involvement, issue involvement, healthy products.

1. Introduction

An important way for fabricants to increase purchase intention of consumers regarding their

products is through differentiation. For the fabricants it is a continuous challenge to distinguish their products and in order to do so the product’s attributes are a useful tool.

A product can be defined by its attributes and consist of three types: features, functions and benefits. This means that a feature will create a certain function, which will turn into a benefit (Crawford & Di Benedetto, 2008). Today’s society is highly competitive and therefore many

(4)

companies try to make use of attributes to differentiate their products. According to Aaker (2003) an attribute will be differentiating when it offers consumers a meaningful proposition. Product

attributes then can turn into a competitive advantage and convince consumers to buy the product over a competing product (Shimp, 2007). An example of a product’s attribute that is differentiating is the “Puur&Eerlijk”-logo used by Albert Heijn. Puur&Eerlijk means that products from this assortment are produced, grown or procured with extra care for human, animal, nature and the environment (“AH puur&eerlijk”, 2013). Consumers that want to purchase healthy and sustainable food are most likely influenced by this product attribute.

When comparing certain attributes, the obtained information can be relevant or irrelevant to the product’s performance (Meyvis & Janiszewski, 2002; Brown & Carpenter, 2000). A relevant attribute describes the main features of a certain product and will highly influence the consumer’s purchase intent. An irrelevant attribute appears to make a significant difference, but in fact is insignificant for the performance of the product (Brown & Carpenter, 2000). Carpenter et al. (1994) argue that an irrelevant attribute changes the decision making process a consumer faces and that it adds value to a meaningless differentiation. In their study they also pointed out that a product with an irrelevant attribute has a more positive impact on product choice than the same product without the irrelevant attribute.

An attribute can be irrelevant, or trivial, but still have an impact on a consumer’s purchase decision. In other words, consumer can be misled by product attributes and purchase a product because of other reasons then they are conscious about at that moment. Product attributes can therefore be deceptive to the consumer. The example provided earlier concerning Albert Heijn’s Puur&Eerlijk-logo is also accused of being deceptive towards the consumers. Recently, there is been a lot of criticism on the using of the logo (Opmeer, 2013; “AH puur&eerlijk”, 2013; Moorman, 2013). The logo creates an image of being healthy and suggests that it does not contain any supplements or factory adjustments. However, with lots of these products this image is false and products are likely

(5)

to have undergone chemical processes (Moorman, 2013). According to Foodwatch, the Dutch food-watchdog, lots of products carrying the Puur&Eerlijk-logo contain the superfluous ingredient sugar. Amongst these products are tomato crème soup, pots with peas or beans, red cabbage with apples, pangasius filet, chicken breast and burgers made from free-range meat. This meat is even filled up with water and contains only about 80% meat (“AH puur&eerlijk”, 2013). Another example of a deceptive attribute used by Albert Heijn is the conscious choice label which indicates a more healthy or better choice within a certain product category. However, in practice this means that the product is simply the least unhealthy in an unhealthy product category and has nothing to do with health in an absolute sense (Moorman, 2013). In the last week of January 2014, the well-known Dutch TV-show Radar, which is about consumer problems and watched by over two million people (SKO Kijkcijfers, 2014), dedicated their broadcast to the deceptiveness of product packages. This broadcast emphasizes the actuality of the problem and began with the statement that “Manufacturers do a lot to mislead you to purchase a product” (Hertsenberg & Buizer, 2014).

The product category regarding healthy food and beverages is highly susceptible to this kind of attributes. In today’s society people get increasingly aware of the importance of good health. Health is a subject that influences and occupies people daily. Research has indicated how severe the consequences of being unhealthy are. These consequences include premature death, high healthcare costs and disabilities (Sun, 2008), the increase of chronic diseases and obesity (Roininen & Tuorila, 1999; Story, Kaphingst, Robinson-O'Brien, & Glanz, 2008); (Stevenson, Doherty, Barnett, Muldoon, & Trew, 2007; Story et al., 2008) and especially child obesity (Gelperowic & Beharrell, 1994; Stevenson et al., 2007). From another perspective, being healthy has a large contribution to a person’s well-being. Studies reveal a strong relationship between health and happiness. Being healthy is linked to higher levels of happiness, and for instance diseases have negative and possibly lasting effects on happiness. Therefore, health is one the most influential factors of well-being (Graham, 2008).

(6)

Despite the increasing awareness of the importance of consuming healthy food, there is still an alarming shortage of knowledge about the relationship between food and health. This results in statements made by food scientists, that at inspection appear to be lacking proof or are simply not true. Scientific research is constantly coming up with new findings changing current health claims. Thereby, food manufacturers are not obliged to mention what ingredients their product contains. Manufacturers often hide certain information about their product. There are lots of ways to conceal unsympathetic sounding substances or hide them with meaningless terms (Moorman, 2013).

Therefore, it is important for consumers to be aware of these risks and that are able to make healthy choices on their own. Biltekoff (2010) adds that people’s beliefs and behaviours are very complex when it comes to consuming healthy food. Eating habits are often not in accordance with the scientific ideals explaining how foods affect bodily functions. People’s aspirations, values and social context are important for establishing a certain diet. In other words, how involved people are with healthy food.

The level of involvement of a consumer can differ on many aspects. For instance, the time they spend on making a purchase decision. The consumer can buy the first product at hand or thoroughly investigate the product’s package, look at the attributes, read the ingredients and compare it to similar products. In order to know what to look for and what all the information present on a product’s package, consumer need to be knowledgeable about what is healthy and what is not. Doing research online, reading books and magazines, consulting experts or friends and family are all ways to increase this knowledge. This requires a higher level of involvement and is important because consumers can be misled by their own interpretation of health food and also the awareness of the possible deceptive factors in advertisement. This level of involvement can be entirely submissive to the advertisements’ message or the products’ attributes and take the

assumption of being healthy for granted. Consumers who are more involved in the process of buying health food can still be misled by the producers. These consumers will be aware of the deceptive factors and look past certain attributes to the actual ingredients for instance. However, a product

(7)

containing little calories, little fat and little sugar can still be not good for a person’s health. Actually, research has shown that most consumers think they eat healthy food, but in reality the opposite is true (Van Dillen, Hiddink, Koelen, de Graaf, & van Woerkum, 2003).

Research has already been done to investigate the effects of deceptive attributes (Brown & Carpenter, 2000; Carpenter, Glazer, & Nakamoto, 1994; Gardner, 1975; Jacoby & Small, 1975). The relationship between deceptive attributes and purchase intention is also mentioned in existing literature (Ailawadi & Neslin, 1998; Grewal, Krishnan, Baker, & Borin, 1998; Halford, Gillespie, Brown, Pontin, & Dovey, 2004). The effect of involvement on this relationship however, is not been

thoroughly researched before. Also, the linkage of these relationships to healthy products cannot be found in existing literature. Since product attributes are constantly changing and new products are introduced to the consumers, more research is necessary to understand all its effects. The aim of this study is therefore to research the influence of involvement on the relationship between deceptive attributes and purchase intention. The involvement on this relationship is divided in product involvement and issue involvement. This leads to the following research question:

What is the effect of product and issue involvement on the relationship between deceptive attributes and the purchase intention of healthy products?

This study contributes to the knowledge of existing academic research in three ways. First of all, an addition is provided to the literature about irrelevant attributes. This study provides more insights about the effects these deceptive attributes have on purchase intention. Beside this, the overall effect of irrelevant attributes on purchase intention is linked to the purchase intention of healthy food specifically. Therefore, this study also aims to make a contribution to the awareness of the consumers about the level of healthiness of the products available. Research is very lacking on the effects of certain products. Above all, this research investigates the effect of involvement on the relationship between irrelevant attributes and purchase intention healthy products. The contribution here is considerable, because there is not a lot of research done in this field. Involvement can play a

(8)

large role in a consumer’s decision making and a consumers is rarely aware of the significance of this role.

In the following chapter, the theoretical framework is discussed in the literature review and leads to the introduction of the hypotheses. Then, the methodology and data collection are

described. The findings from the data collections and the answering of the hypotheses will be presented in the results part. This paper will end with the discussion and conclusion, followed by the limitations and the recommendations for future research.

(9)

2. Literature review

2.1 The effects of product attributes on purchase intention

A product consists entirely of attributes, and any product can be described by numerating its attributes (Crawford & Di Benedetto, 2008). Product attributes are a main influencer on consumer decision making and thus purchase intention. Existing literature contains research into the processing of product attributes, purchase intention and actual consumption (Ailawadi & Neslin, 1998; Saffer & Dave, 2006; Halford et al., 2004; Coon, Goldberg, Rogers, & Tucker, 2001).

In order to evaluate the purchase intention of health food, there is a need to analyse the product attribute processing and the decision making process. One of the components researched in existing literature is the interaction between memory and the decision process. Suggested is that there are both memory and stimuli aspects and these two interact when deciding to buy or consume a product. Interesting is the role of expertise when making decisions, as well as inferences and prior knowledge (Bettman, 1986).

Next to the persuasiveness of advertisements in general, consumer memory, and especially the retrieval cues, for these advertisements is very important. These factors also have the same significance concerning the persuasiveness of product attributes. This is because consumers do not make brand decisions when they are exposed to advertisements or product attributes. For instance a cue like a certain ingredient, will not retrieve thoughts, feelings or information from a person’s memory stored there from a previous experience. Verbal or visual information can facilitate the memory retrieval process. Other factors affecting consumer advertisement memory and brand evaluations are the number of brands advertising within a certain category and the processing goals of the consumer during the advertisement exposure. Especially the placement of retrieval cues on the product or package can have a major effect on a consumer’s decision. These cues should increase the likelihood that consumers retrieve contents of the advertisement and the product’s attributes. The memory trace consist of learned, heard, felt or seen information by the consumer when exposed

(10)

to an advertisement. In other words, these cues should, after the retrieval of elements, lead to a more favourable evaluation and higher likelihood of purchase of the product (Keller, 1987).

(Grewal et al., 1998) also state that consumers use certain cues as signals for price discounts, store name and brand name. Retailers or producers that understand how to use the influence of these cues can influence the decision making process and improve their competitive situation. Later research between online and traditional supermarkets underlines this finding by saying that

consumers use prior information already available when making decisions as well as information they can obtain from their external environment. Consumers will focus on relevant attributes that are diagnostic and available in the external environment (Degeratu, Rangaswamy, & Wu, 2000).

Highlighting your product’s attributes is a successful way in creating cues and steering the decision-making process. This way you can also differentiate your product from others. However, you can also be successful in differentiating by using attributes that appear valuable, but are actually irrelevant to creating a certain benefit. People tend to value these differentiating attributes despite their irrelevance (Carpenter et al., 1994). In other words, consumers then face some sort of

deception. Presently, there is an increasing intention of these deceptive product attributes on healthy food consumption. This market will be discussed in the next paragraph, followed by the elaboration on deceptive product attributes.

2.2 The effects of healthy and unhealthy food consumption

A lot of research has been done to investigate the effects of consuming unhealthy food and consequently the emergence of need for health food. The negative outcomes of consuming

unhealthy food are clearly mentioned in a wide range of existing literature. For instance, in the USA a poor diet is estimated to cause up to 580.000 deaths each year. To decrease this death toll, special taxes for unhealthy food and subsidies for healthy food is suggested to create a healthier

environment. The potential income from the taxes could be reinvested in means to improve the 10

(11)

consumption of health food (Jacobson & Brownell, 2000). Obesity is probably the best known problem in society regarding food consumption. Obesity is seen as a major consequence of unhealthy food, especially among children (Gelperowic & Beharrell, 1994).

Health is seen as a consequent factor determining one’s own happiness and it even surpasses common belief systems and national boundaries. Economists researching happiness developed equations to formulate happiness and heath is seen as one of the variables causing variation in one’s happiness. Even more, health is described as the variable that correlates most strongly with happiness. Even more than a person’s income (Graham, 2008).

Both perspectives, consumption of healthy food and unhealthy food, still intrigues a lot off researchers. Especially the topic of why consumers remain to consume unhealthy food despite the negative consequences. A possible explanation could be that people are simply still not aware of the detrimental outcomes, which could be caused by a lack of good marketing or of the presentation of nutritional information. Another explanation is that people lack the self-will to resist unhealthy food. This could be caused by the perception of the consumer that unhealthy food tastes better than healthy food (Raghunathan, Naylor, & Hoyer, 2006). More understanding about the available products in the market is necessary and especially on how to advertise these products in order to help consumers make better choices in deciding about their food consumption.

Moreover, food choice is seen as a process that contains social, economic, psychological, cultural and biological factors. These factors are interacting with the experiences someone encounters during their lifetime and will eventually turn into personal preferences like taste and considerations like convenience. Besides this, certain identities are related to food consumption based on everyday interaction with groups, objects and people. These identities reflect layers of classification that are structural, cultural, individual and social. Identities are form through particular traits of a person, the relationship with collective group categories and the social role of a person (Bisogni, Connors, Devine, & Sobal, 2002).

(12)

Gelperowic and Beharrell (1994) describe the negative consequences of consuming

unhealthy food, where Raghunathan (2006) claims this could be caused by perception. It is obviously a complex issue why people choose to consume a particular food and therefor marketing can play a significant role in forming and steering food consumption. In other words, marketing, and more specifically marketing of product attributes, can influence the purchase intention of the consumer. Especially, if there is any form of deceptiveness that originates from these product attributes.

2.3 Deceptive product attributes

Besides the actual processing of product attributes, it is important to take into account that these attributes can be deceptive and potentially contain misleading information. Aditya (2001) describes deception in a marketing environment as something that is perceived as unfair or unethical. In essence, deception is seen as an act that is misleading a certain target group. In addition, Gardner (1975) explains that deception occurs when a consumer is provided with a belief or impression that is different from what could reasonably be expected and that is possibly misleading or factually false.

According to Carpenter et al. (1994) the adding of an irrelevant attribute to a product changes the structure of the decision making process the consumers deal with. Even more when the attribute is difficult to evaluate. The consumer is likely to conclude that the attribute is actually valuable. This positive judgement of an irrelevant attribute depends on the awareness of the

consumer about the true meaning of this attribute. The study shows that consumers rated a product more positively with an irrelevant attribute than the same product without that irrelevant attribute. This is partly due to the differentiating effect the irrelevant attributes have compared to other products. In line with these findings, Mukherjee and Hoyer (2001) state that the improvement of product evaluation can be caused by the addition of a new and unique product attribute.

These studies show a positive effect of irrelevant attributes on product evaluation. However, there are also studies that did not found any positive relations between irrelevant attributes and the 12

(13)

way consumers rate a product. In other words, the effect that irrelevant attributes have on purchase intention. The study of Meyvis and Janiszewski (2002) found that consumers search for information on product attributes in order to evaluate its value. When they encounter information that is perceived as irrelevant, their beliefs about the product will be weakened (Meyvis & Janiszewski, 2002; Burke, 2006). In addition, Simonson, Carmon and O’Curry (1994) reason that an attribute is less attractive when it has no contribution to the value of the product. The product’s attribute may have a negative effect on a brand’s overall choice probability.

Concerning these findings, one can assume that in a continuously changing and increasingly competitive environment people face more and more products and their attributes every day. Consumers are more critical about their purchases and seem to attach more value to trustworthy brands and ethicality. Therefore, the following hypothesis is formulated:

H1: Deceptive product attributes have a negative effect on purchase intention.

An example of consumer behaviour influenced by product attributes is given in the study by Gelperowic (1994). This study discovered that parents, as the actual customers, care more about the food consumption of their children. The influence of children in the purchase process is significant. This influence is formed by the preference of children, which is often affected by packaging, shapes and flavours of food items. The study aims at investigating whether the product and packaging factors affect purchase decisions by a mother and her child. The focus is on healthy food products or products perceived as healthy. The influence of children shows that the most appealing products are not always healthy products. Therefore, mothers must have a certain level of involvement in order to recognize the healthy products. This particular level of involvement exists because of parental responsibilities. However, there are more reasons why consumers have a certain degree of involvement and the effect involvement can have on purchase intention.

(14)

2.4 The role of involvement

Research mentions involvement as a potential moderator of the generation of deceptive inferences from advertisements. Time and invalid inferences are included by Johar (1995), as well as purchase intentions. Attributes can be objectively determined to be false. For instance, when claiming to contain an ingredient which is not one of the ingredients of the product. However, it is difficult to inhibit packages or advertisements that contain perfect information from a regulatory standpoint, but still create a misleading impression on the consumer (Jacoby & Small, 1975). This is specifically relevant to the food market, since a regulatory framework is absent where boundaries are set to distinguish between what is health food and what is not. For instance, there are practically no rules about the placement and visibility of ingredients on the package of a product. The only existing rule is that they simply have to be on the package. Because of this, fabricants can use uncountable manners to hide this information from the consumers.

The role of involvement in the relationship between product attributes and purchase intention is an important one. In existing literature it is stated that information about health food is taken in differently by consumers in terms of following the advice regarding health food or not. This distinction is important for the development of promotional strategies for consuming health food (Roininen & Tuorila, 1999). This could also imply that involvement, in terms of knowledge uptake, is a factor that influences the relationship between the information received and the actual purchase intention. Consumers value the Internet as the preferred medium to obtain health-related

information (Williams, Nicholas, Huntington, & McLean, 2002). In addition to this, the family doctor, the social environment, magazines, the Internet, dieticians, television, the Nutrition Centre, food labels, the media and food manufacturers are mentioned as information sources (Van Dillen et al., 2003).

Existing literature about decision processes highlights the effect of involvement. Product involvement is important in understanding consumer attitudes and behaviour because involvement

(15)

is a central motivation factor that shapes the purchase decision-making process. It shapes consumer attitudes and behaviours (Kim, Damhorst, & Lee, 2002). One of the leading academic journals on this subject by Petty, Cacioppo and Schumann (1983) states that involvement is about the existence of central and peripheral routes when processing advertisements. The central route investigates any change in attitude when a person puts in effort when considering certain information. The peripheral route emphasizes change in attitude when the information is associated with positive or negative cues. In other words, they investigate the differences between high and low involvement (Petty, Cacioppo, & Schumann, 1983). High involvement is necessary to make a consumer change his or her eating habits. Furthermore, the message should be processed in detail and the content of the message should be more persuasive when negatively framed. In other words, negative information outweighs positive information (Maheswaran & Meyers-Levy, 1990).

Further research includes factors such as pleasure value, symbolic value, risk importance and risk probability identifying involvement as the level of perceived personal importance, interest or relevance which is evoked by a certain stimulus. This stimulus is then linked by the consumer to specific goals. A function of consumer involvement is their motivational force, which can be an explanation for certain steps in the decision-making process (Verbeke & Vackier, 2004).

Research into involvement is often focused on pre-purchase settings in the decision making process. Little attention is given to the effect of involvement on consumer behaviour on an enduring basis. The existence of different forms of involvement, namely product involvement, purchase decision involvement, advertising involvement and consumption involvement, can form an overall profile of consumer involvement (O’Cass, 2000). Food involvement, introduced by Eertmans (2005), is indicated as an important variable in food choice. Food involvement is the level of importance of food in a person’s life and describes the extent to which people can make discriminations between food items. Van Dillen (2003) describes more educated people and specifically women as being more

(16)

aware of nutrition. Also, young people tend to be more aware of their food consumption because of a change in lifestyle and perceptions of health in general.

This research will focus on two types of consumer involvement. Namely, consumer product involvement and consumer issue involvement. Product involvement is the level of a consumer’s awareness of product attributes and interest in purchasing a certain product. Also, the willingness to do extensive research in order to increase knowledge regarding a product type. This leads to the following hypothesis:

H2: Consumer product involvement moderates the effect of deceptive product attributes on purchase intention. The effect of deceptive product attributes is less negative when there is a higher level of consumer product involvement.

Issue involvement describes the extent to which the issue of deceptive attributes on health products is of personal importance. According to Petty & Cacioppo (1979), high issue involvement occurs when an issue has personal meaning or intrinsic importance. It can also occur when people expect the issue to have significant consequences for their own lives. This leads to the following hypothesis:

H3: Consumer issue involvement moderates the effect of deceptive product attributes on purchase intention. The effect of deceptive product attributes is less negative when there is a higher level of consumer issue involvement.

The variables covered by this research, the relationships between these variables and the hypotheses concerning these variables are shown in the conceptual model in figure 1 on the next page.

(17)

Figure 1: Conceptual Model

(18)

3. Data and method

In this chapter, the data collection and the research method will be described. In order to answer the research question and the hypotheses presented in the previous chapters, an online survey was conducted. The survey was conducted in Dutch, because all the respondents are living in the Netherlands. The respondents were asked to give their opinion about the products shown in the survey. The product-type is a strawberry juice, because it is a familiar product-type and easily related to the different variables this research investigates. The respondents where shown four different products from the same product type, namely Product A, Product B, Product C and Product D. These names are chosen to indicate the difference between the products, without causing any dissimilarity in attitudes towards the brand name. The image of the products are followed by four identical questions related to the variables used in this research. The difference between the four products was whether they claimed to be healthy or not and if they actually where healthy or not. Product A claimed to be healthy and the product attribute was actually healthy, because it shows the “healthy choice”-label and the attribute “without concentrates” (see figure 2).

Figure 2: Product A

Product B claimed to be healthy but the product attribute was in fact not healthy, because it shows the “healthy choice”-label and the attribute “with natural sugars” (see figure 3). However, the

(19)

addition of natural sugars are most likely fruit sugar concentrates, which are still sugars and equally bad for one’s health.

Figure 3: Product B

Product C did not claim to be healthy but the product attribute was healthy, because it shows no “healthy choice”-label and the attribute “no extra sweeteners” (see figure 4).

Figure 4: Product C

(20)

Product D did not claim to be healthy and the product attribute was not healthy, because it shows no “healthy choice”-label and the attribute “extra fruit filling” (see figure 5). Extra fruit filling is very likely to involve honey-sweet fruit concentrates with a noxiousness equal to pure sugar.

Figure 5: Product D

The first part of this chapter elaborates on the sample used, which is followed by the reliability explanation for the variables “issue involvement”, “product involvement”, “misleading product attributes” and “purchase intention”.

3.1 Sample

The data collection was done by asking respondents to fill in a questionnaire created using Qualtrics. 204 respondents participated in this study of which 167 fully completed the survey. The survey was conducted online by sharing a link on social media platform Facebook. Also the author send the link through email to some friends and relatives. In this email the request was made to forward to link anyone willing to help or interested. This way the number of respondents with different backgrounds was increased. From these participants 52% were male. The average age of the respondents was 40.5 years (range 18-82, SD 18.1). A majority of the respondents (84%) have finished or are currently

(21)

following a higher education at HBO/WO (bachelor/master) level. The remaining respondents possess a HAVO/VWO/Gymnasium high school diploma.

3.2 Reliability

3.2.1 Issue involvement

Originally the scale for issue involvement consisted of 20 items. The reliability analysis, presented in table 1 and 2, shows that even though Cronbach’s alpha for this scale is well above the lower boundary of .60, the first four items correlate poorly with the rest of the scale. These items are the questions “How healthy do you think Product A is?”, “How healthy do you think Product B is?”, “How healthy do you think Product C is?” and “How healthy do you think Product D is?”. Respondents could answer a 7-points bipolar scale ranging from “unhealthy” till “healthy”.

Table 1: Reliability Statistics

Cronbach's Alpha N of Items

,859 20

(22)

Table 2: Item-Total Statistics Scale Mean if Item Deleted Scale Variance if Item

Deleted Total Correlation Corrected

Item-Cronbach's Alpha if Item

Deleted How healthy do you think Product A

is?-Unhealthy:Healthy 83,00 330,552 ,001 ,868

How healthy do you think Product B

How healthy do you think Product C

How healthy do you think Product D

I am interested in buying healthy

products.-Uninterested:Interested 81,43 305,599 ,564 ,850

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Nitrite

83,68 281,751 ,591 ,846

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Aspartame

82,69 289,704 ,477 ,852

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Dyes

82,25 284,173 ,677 ,842

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Fructose

82,85 288,432 ,591 ,846

To me, healthy eating and drinking is:-Not

important:Important 81,19 308,197 ,567 ,850

To me, healthy eating and drinking

is:-Unaffordable:Affordable 81,85 316,529 ,283 ,858

To me, healthy eating and drinking is:-Never on my

mind:Always on my mind 81,87 305,948 ,537 ,850

is:-Uninteresting:Interesting 81,31 310,311 ,490 ,852

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I do research online

83,47 297,161 ,470 ,852

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I read books or magazines about health

83,20 290,574 ,602 ,846

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I consult experts

84,41 302,492 ,454 ,852

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I ask family and friends

82,99 304,000 ,430 ,853

Are you aware of the following dangers? Hereby, 1 is

“Not aware” and 7 is “Aware”.-Insulin resistance 83,68 286,358 ,598 ,846

“Not aware” and 7 is “Aware”.-Saturated fats 81,85 294,625 ,603 ,846

Are you aware of the following dangers? Hereby, 1 is “Not aware” and 7 is “Aware”.-Harmful chemical substances

82,57 285,199 ,678 ,842

Table 3 and 4 show that after removing the four items mentioned above, alpha increases to .89 and all items have positive item-total correlations. Alpha cannot be improved any further so we will keep these 16 items to calculate the respondents’ average issue involvement scores.

(23)

,893 16

Table 4: Item-Total Statistics

Scale Mean if Item Deleted

Scale Variance if Item

Deleted I am interested in buying healthy

products.-Uninterested:Interested 66,61 293,339 ,615 ,886

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Nitrite

68,83 271,979 ,600 ,886

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Aspartame

67,87 278,393 ,504 ,890

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Dyes

67,41 274,929 ,681 ,882

Are you aware of the effect of the following ingredients? Hereby, 1 is “Not aware” and 7 is “Aware”. -Fructose

68,07 275,647 ,642 ,883

To me, healthy eating and drinking is:-Not

important:Important 66,35 298,121 ,585 ,888

is:-Unaffordable:Affordable 67,00 308,470 ,250 ,896

To me, healthy eating and drinking is:-Never on my

mind:Always on my mind 67,06 293,560 ,601 ,886

is:-Uninteresting:Interesting 66,48 298,023 ,557 ,888

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I do research online

68,69 287,089 ,480 ,890

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I read books or magazines about health

68,42 278,769 ,638 ,884

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I consult experts

69,61 290,267 ,505 ,889

What do you do to find information about health? Hereby, 1 is “Never” and 7 is “Always”.-I ask family and friends

68,18 295,558 ,411 ,892

“Not aware” and 7 is “Aware”.-Insulin resistance 68,85 275,052 ,625 ,884

“Not aware” and 7 is “Aware”.-Saturated fats 67,01 284,664 ,621 ,885

Are you aware of the following dangers? Hereby, 1 is “Not aware” and 7 is “Aware”.-Harmful chemical substances

67,76 275,284 ,682 ,882

3.2.2 Product involvement

The 22-item product involvement scale seems to have a satisfactory alpha at first sight (alpha = .88), but after inspecting the item-total correlations we see that the five items with the best content

(24)

formulation regarding product involvement, correlate poorly with the rest of the scale (Table 5 and 6). These five items are “How attentively do you look at the product’s attributes?”, “I inspects the product’s package”, “I look for the product’s attributes”, “I read the list of ingredients” and “I am aware of whether a product is healthy or not”. The other 17 items seem to have more to do with product evaluation. However, product evaluation would be an alternative dependent variable next to purchase intention and represents a different scale. Therefore, it will not be included in this research.

,879 22

Deleted

I experience Product C as:-Not credible:Credible 80,46 291,750 ,627 ,869

I experience Product C as:-Bad:Good 80,65 294,395 ,656 ,869

I experience Product C as:-Unattractive:Attractive 80,89 295,065 ,614 ,870

I experience Product C as:-Misleading:Not misleading 80,58 297,170 ,558 ,872

I experience Product D as:-Not credible:Credible 81,12 294,076 ,613 ,870

I experience Product D as:-Bad:Good 80,91 294,810 ,620 ,870

I experience Product D as:-Unattractive:Attractive 81,29 297,345 ,571 ,871

I experience Product D as:-Misleading:Not misleading 80,79 299,789 ,516 ,873

How attentively do you look at the product’s

attributes?-Not attentive:Attentive 79,33 331,541 -,085 ,890

I have confidence in the products that are for sale in

the supermarket.-Disagree: 80,47 320,372 ,119 ,884

What do you do to find information about healthy products? Hereby, 1 is “Never” and 7 is “Always-I inspect the product’s package

79,60 318,408 ,126 ,886

What do you do to find information about healthy products? Hereby, 1 is “Never” and 7 is “Always.-I look for the product’s attributes

79,92 320,434 ,096 ,886

What do you do to find information about healthy products? Hereby, 1 is “Never” and 7 is “Always.-I read the list of ingredients

80,23 324,918 ,006 ,891

I am aware of whether a product is healthy or

not.-Unaware:Aware 78,95 320,149 ,195 ,881

I experience Product A as:-Not credible:Credible 80,44 286,082 ,715 ,866

I experience Product A as:-Bad:Good 80,23 288,407 ,772 ,865

I experience Product A as:-Unattractive:Attractive 80,79 292,107 ,662 ,868

I experience Product A as:-Misleading:Not misleading 80,22 301,460 ,482 ,874

I experience Product B as:-Not credible:Credible 80,49 289,843 ,706 ,867

I experience Product B as:-Bad:Good 80,62 291,375 ,740 ,867

I experience Product B as:-Unattractive:Attractive 80,74 291,362 ,699 ,867

I experience Product B as:-Misleading:Not misleading 80,48 295,464 ,598 ,870

(25)

Table 7 and 8 show that after removing the 17 product evaluation items, there is a reliable 5-item scale for product involvement (alpha=.86).

,857 5

Deleted How attentively do you look at the product’s

attributes?-Not attentive:Attentive 18,77 29,766 ,619 ,840

What do you do to find information about healthy products? Hereby, 1 is “Never” and 7 is “Always-I inspects the product’s package

19,07 27,415 ,693 ,822

What do you do to find information about healthy products? Hereby, 1 is “Never” and 7 is “Always.-I look for the product’s attributes

19,39 26,080 ,794 ,793

What do you do to find information about healthy products? Hereby, 1 is “Never” and 7 is “Always.-I read the list of ingredients

19,60 24,614 ,756 ,806

I am aware of whether a product is healthy or

not.-Unaware:Aware 18,40 34,548 ,536 ,861

3.2.3 Misleading product attributes

The next scale measures the respondents’ sensitivity to misleading product attributes; high scores on this scale reflect respondents that attach relatively much value to misleading product attributes. Table 9 presents that the scale is reliable (alpha = .79) and table 10 shows that all item-total correlations are positive. It must be noted that the second item of this scale (“I was aware of the presence of misleading attributes with the products from the last questions”, ranging from “Not aware” till “aware”) is reversed, because lower scores reflect unawareness of the misleading product attributes, which may be associated with higher sensitivity to misleading product attributes.

,788 14

(26)

Scale Mean if

Item Deleted if Item Deleted Scale Variance Total Correlation Corrected Item- Cronbach's Alpha if Item Deleted Indicate to what extent Product D pleases

you:-Does not please:Pleases 48,5658 143,082 ,306 ,783

I was aware of the presence of misleading attributes with the products from the last questions.-Not aware:Aware

49,1513 147,546 ,224 ,788

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"No sugar added"

47,0329 136,111 ,413 ,774

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"With natural sugars"

48,0592 132,983 ,506 ,766

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"Good for heart and vanes"

47,5461 134,620 ,490 ,767

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"Extra fruit filling"

48,3882 136,570 ,422 ,773

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"With extra fibres"

47,5921 135,925 ,525 ,765

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"Less calories"

47,7171 135,529 ,475 ,769

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"Ik kies bewust"

47,5526 138,938 ,376 ,777

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"Puur & Eerlijk"

47,6842 139,317 ,363 ,779

How do you estimate the value of the following product attributes? Hereby, 1 is “Not valuable” and 7 is “Valuable”-"Light"

48,4671 140,926 ,295 ,785

Indicate to what extent Product A pleases

Indicate to what extent Product B pleases

Indicate to what extent Product C pleases

3.2.4 Purchase intention

Table 11 presents the initial internal consistency of the purchase intention scale which is low, alpha=.53. Inspecting the content and item-total correlations of the items, we see low item-total correlations at items that do not directly assess purchase intention, as can be seen in table 12. These items are “How much are you willing to pay more for healthy products?”, “How much do you spend on healthy products per week?”, “How often do you intent to purchase healthy products before going to the store?”, “How often do you purchase healthy products without this intention

(27)

beforehand?”, “Products in the supermarket are trying to convince me to purchase them” and “Products in the supermarket are putting me under pressure to purchase them”.

,530 10

Scale Mean if

Item Deleted Scale Variance if Item Deleted Total Correlation Corrected Item- Cronbach's Alpha if Item Deleted How likely are you to purchase Product

A?-Unlikely:Likely 32,31 33,361 ,434 ,432

How likely are you to purchase Product

B?-Unlikely:Likely 32,35 34,004 ,456 ,431

C?-Unlikely:Likely 32,46 34,132 ,444 ,434

D?-Unlikely:Likely 32,83 37,603 ,256 ,495

How much are you willing to pay more for

healthy products? 32,94 42,306 ,085 ,536

How much do you spend on healthy products

per week? 32,32 42,245 ,022 ,559

How often do you intent to purchase healthy products before going to the

store?-Never:Always

30,24 42,207 ,047 ,549

How often do you purchase healthy products without this intention

beforehand?-Never:Always

31,41 40,519 ,107 ,537

Products in the supermarket are trying to convince me to purchase

them.-Disagree:Agree

30,73 37,882 ,190 ,516

Products in the supermarket are putting me under pressure to purchase

them.-Disagree:Agree

32,71 38,551 ,213 ,507

Table 13 and 14 show that after removing these items, the purchase intention scale is highly reliable (alpha=.81).

,810 4

Scale Mean if

Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted How likely are you to purchase

Product A?-Unlikely:Likely 8,99 14,522 ,674 ,739

How likely are you to purchase

Product B?-Unlikely:Likely 8,99 14,963 ,728 ,714

Product C?-Unlikely:Likely 9,14 15,832 ,629 ,761

Product D?-Unlikely:Likely 9,48 17,506 ,490 ,823

(28)

4. Results

In this chapter, the findings from the data collection will be discussed. The information derived from the data collection is analysed and the hypotheses will be rejected or accepted. First, the descriptives and correlations are discussed and subsequently the conceptual model is tested.

4.1 Descriptives and correlations

The descriptive statistics of the scale scores are shown in table 15 below. The average product involvement score is relatively high; 4.8 on a 7-point scale. This indicates that the respondents are paying good attention to what they buy. The average score on issue involvement (4.5) is also above the “neutral” score of 4; this indicates that the respondents pay attention to health aspects. Purchase intention is relatively low (3.1); the respondents do not feel compelled to buy the advertised products. The sensitivity to misleading product attributes is also a little below the “neutral” score of 4.

Table 15: Descriptive Statistics

N Minimum Maximum Mean Std. Deviation

Issue involvement 160 1,56 7,00 4,4925 1,11717

Product involvement 161 1,60 7,00 4,8069 1,31099

Misleading product attributes 171 1,00 6,07 3,6462 ,94379

Purchase intention 171 1,00 7,00 3,0741 1,34582

Valid N (listwise) 159

Table 16 shows the scale score when the variables misleading product attributes and purchase intention are measured per product. Misleading product attributes is now formulated as the level of awareness of the consumer about misleading attributes.

The average product involvement and issue involvement scores remain the same. Purchase intention ranges from 2.7 to 3.2 for the different products. In other words, the respondents do not feel compelled to buy any of the advertised products. The awareness of misleading product

attributes ranges from 3.4 to 3.8 for the different products, which is below the “neutral” score of 4.

(29)

Table 16: Descriptive Statistics

N Minimum Maximum Mean Std. Deviation

Issue involvement 160 1,56 7,00 4,4925 1,11717

Product involvement 161 1,60 7,00 4,8069 1,31099

I experience Product A as:-Misleading:Not misleading 164 1 7 3,80 1,591

I experience Product B as:-Misleading:Not misleading 161 1 7 3,65 1,551

I experience Product C as:-Misleading:Not misleading 157 1 7 3,57 1,574

I experience Product D as:-Misleading:Not misleading 158 1 7 3,35 1,518

How likely are you to purchase Product A?-Unlikely:Likely 171 1 7 3,22 1,741

How likely are you to purchase Product B?-Unlikely:Likely 167 1 7 3,19 1,574

How likely are you to purchase Product C?-Unlikely:Likely 165 1 7 3,07 1,608

How likely are you to purchase Product D?-Unlikely:Likely 165 1 7 2,74 1,577

Valid N (listwise) 142

The correlation matrix presented in table 17 demonstrates a significant correlation between product involvement and issue involvement, r=.74, p<.001. This indicates that people that are more

preoccupied with health issues, also pay more attention to what they buy.

Furthermore there is a significant correlation between purchase intention and sensitivity to misleading product attributes, r=.59, p<.001. This indicates that people that are more sensitive to misleading product attributes have higher purchase intentions regarding the stimuli products.

Table 17: Correlations

Issue involvement involvement Product Misleading product attributes intention Purchase

Issue involvement Pearson

Correlation 1 ,744

** _-,047 _-,121

Sig. (2-tailed) ,000 ,557 ,126

N 160 159 159 160

Product involvement Pearson

Correlation ,744

** ₁ _-,125 _-,198*

Sig. (2-tailed) ,000 ,114 ,012

N 159 161 161 161

Misleading product

attributes Pearson Correlation -,047 -,125 1 ,587

**

Sig. (2-tailed) ,557 ,114 ,000

N 159 161 171 170

Purchase intention Pearson

Correlation -,121 -,198

* _,587** ₁

Sig. (2-tailed) ,126 ,012 ,000

N 160 161 170 171

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

(30)

The correlation matrix in table 18 demonstrates the correlations as measured in table 17, but now also per product separately. The four different purchase intention (PI) outcomes and the four

different awareness of misleading product attributes (mpa) outcomes correlate significantly (2-tailed) among each other.

Furthermore, there are significant (2-tailed) positive correlations between the purchase intention items and the awareness of misleading product attributes. This indicates that people that are “more” less aware of misleading product attributes (higher scores indicate less awareness) have higher purchase intentions (higher scores indicate more purchase intention) regarding the stimuli products.

Table 18: Correlations

Issue

involvement involvement Product mpa Pr. Aware A Aware mpa Pr. B Aware mpa Pr. C Aware mpa Pr. D PI Pr. A PI Pr. B PI Pr. C PI Pr. D Issue involvement R 1 ,744** -,268** -,115 -,108 -,213** -,089 -,036 -,079 -,180* Sig. ,000 ,001 ,161 ,184 ,008 ,263 ,654 ,322 ,023 N 160 159 153 151 152 153 160 157 158 159 Pr. involvement R ,744** 1 -,326** -,199* -,265** -,270** -,110 -,161* _,223**- -,202 * Sig. ,000 ,000 ,014 ,001 ,001 ,165 ,043 ,005 ,010 N 159 161 155 153 154 155 161 159 160 161 Aware mpa Pr. A R -,268** -,326** 1 ,567** ,445** ,438** ,548** ,453** ,373** ,365** Sig. ,001 ,000 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 153 155 164 158 156 156 164 161 159 159 Aware mpa Pr. B R -,115 -,199* _,567** ₁ _,461** _,391**_,430**_,647**_,345**_,366** Sig. ,161 ,014 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 151 153 158 161 152 153 161 159 157 157 Aware mpa Pr. C R -,108 -,265** ,445** ,461** 1 ,518** ,458** ,472** ,748** ,356** Sig. ,184 ,001 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 152 154 156 152 157 155 157 155 157 157 Aware mpa Pr. D R -,213** _-,270** _,438** _,391** _,518** _{1 ,298}**_,336**_,386**_,637** Sig. ,008 ,001 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 153 155 156 153 155 158 158 156 157 158 PI Pr. A R -,089 -,110 ,548 ** _,430** _,458** _,298** _{1 ,697}**_,547**_,439** Sig. ,263 ,165 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 160 161 164 161 157 158 171 167 165 165 PI Pr. B R -,036 -,161 * _,453** _,647** _,472** _,336**_,697** _{1 ,610}**_,438** Sig. ,654 ,043 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 157 159 161 159 155 156 167 167 163 163 PI Pr. C R -,079 -,223 ** _,373** _,345** _,748** _,386**_,547**_,610** _{1 ,440}** Sig. ,322 ,005 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 158 160 159 157 157 157 165 163 165 164 PI Pr. D R -,180* -,202* ,365** ,366** ,356** ,637** ,439** ,438** ,440** 1 Sig. ,023 ,010 ,000 ,000 ,000 ,000 ,000 ,000 ,000 N 159 161 159 157 157 158 165 163 164 165

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

(31)

4.2 Model testing

In this paragraph the following model, as described in the literature review and once again presented in figure 6, is tested:

Figure 6: Conceptual Model

We use a linear regression analysis with purchase intention as the dependent variable, gender and age are entered in the first analysis as control variables. These control variables make sure that the results of the analysis are not biased by effects that are attributable to gender or age. This means that possible gender or age effects do not distort the results of the analysis. The coefficients of the other variables are 'controlled for' gender and age. Next, the independent variables issue

involvement, product involvement and misleading product attributes are added. Finally we test for moderator (interaction) effects by entering the product variables of misleading product attributes * issue involvement and misleading product attributes * product involvement. To avoid

multicollinearity problems with the original independent variables and the product variables, the original variables are first standardized before calculating the product variables.

The model summary presented in table 19 below demonstrates no significant effects of the control variables, F(2,153)=1.1, r2_{=1%, ns. The three predictors in the model however do have a}

(32)

significant effect on purchase intention, Fch(3,150)=28.3, r2_{ch=35.6%, p<.001. The extra variance} explained by the moderators is just not significant, Fch(2,148)=2.9, r2_{ch=2.3%, p=.06.}

Table 19: Model Summary

Model R

R Square

Adjusted R

Square Std. Error of the Estimate

Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,119a ,014 ,001 1,32194 ,014 1,106 2 153 ,334 2 ,609b ,371 ,350 1,06680 ,356 28,312 3 150 ,000 3 ,628c ,394 ,365 1,05374 ,023 2,870 2 148 ,060

a. Predictors: (Constant), My gender is:, My age is:

b. Predictors: (Constant), My gender is:, My age is:, Misleading product attributes, Product involvement, Issue involvement

c. Predictors: (Constant), My gender is:, My age is:, Misleading product attributes, Product involvement, Issue involvement, moderator issue involvement, moderator product involvement

d. Dependent Variable: Purchase intention

This analysis is executed four more times, but now once for each of the four stimulus products. The model summaries in tables 20, 21, 22 and 23 also demonstrate no significant effects of the control variables. The three predictors in the model have a significant effect on purchase intention as well. The moderators have a significant effect only with Product C.

A. The effect of the control variables is not significant, F(2,147)=.4, r2_{=.6%, ns.} The effect of the model variables is significant, Fch(3,144)=20.1, r2_{ch=29%, p<.001.} The effect of the moderators is not significant, Fch(2,142)=1.6, r2_ch=1.5%,ns. B. The effect of the control variables is not significant, F(2,144)=.4, r2_{=.1%, ns.}

The effect of the model variables is significant, Fch(3,141)=31.4, r2_{ch=40%, p<.001.} The effect of the moderators is not significant, Fch(2,139)=2.0, r2_{ch=1.7%, p=.06.} C. The effect of the control variables is not significant, F(2,146)=.8, r2_{=1%, ns.}

The effect of the model variables is significant, Fch(3,143)=58.4, r2_{ch=55%, p<.001.} The effect of the moderators is significant, Fch(2,141)=3.7, r2_{ch=2%, p<.05.}

D. The effect of the control variables is not significant, F(2,147)=3.0, r2_{=2.6%, ns.} The effect of the model variables is significant, Fch(3,144)=27.7, r2_{ch=35%, p<.001.} The effect of the moderators is not significant, Fch(2,142)=.3, r2_{ch=.3%, ns.}

(33)

Model R R Square Adjusted R

Square Std. Error of the Estimate _{R Square} Change Statistics

Change F Change df1 df2 Change Sig. F

1 ,077a _,006 _-,008 _1,729 _,006 _,434 _{2 147} _,649

2 ,547b _,299 _,275 _1,467 _,293 _20,085 _{3 144} _,000

3 ,561c ,315 ,281 1,461 ,015 1,603 2 142 ,205

b. Predictors: (Constant), My gender is:, My age is:, I experience Product A as:-Misleading:Not misleading, Product involvement, Issue involvement

c. Predictors: (Constant), My gender is:, My age is:, I experience Product A as:-Misleading:Not misleading, Product involvement, Issue involvement, moderator issue involvement A, moderator product involvement A

1 ,024a ,001 -,013 1,582 ,001 ,041 2 144 ,960

2 ,633b _,401 _,379 _1,238 _,400 _31,385 _{3 141} _,000

3 ,646c ,418 ,388 1,229 ,017 2,009 2 139 ,138

b. Predictors: (Constant), My gender is:, My age is:, I experience Product B as:-Misleading:Not misleading, Product involvement, Issue involvement

c. Predictors: (Constant), My gender is:, My age is:, I experience Product B as:-Misleading:Not misleading, Product involvement, Issue involvement, moderator issue involvement B, moderator product involvement B

1 ,101a _,010 _-,003 _1,615 _,010 _,755 _{2 146} _,472

2 ,745b ,555 ,539 1,094 ,545 58,364 3 143 ,000

3 ,760c _,577 _,556 _1,074 _,022 _3,700 _{2 141} _,027

b. Predictors: (Constant), My gender is:, My age is:, I experience Product C as:-Misleading:Not misleading, Product involvement, Issue involvement

c. Predictors: (Constant), My gender is:, My age is:, I experience Product C as:-Misleading:Not misleading, Product involvement, Issue involvement, moderator issue involvement C, moderator product involvement C

1 ,198a _,039 _,026 _1,573 _,039 _3,001 _{2 147} _,053

2 ,625b ,391 ,369 1,266 ,351 27,676 3 144 ,000

3 ,627c _,393 _,363 _1,272 _,003 _,294 _{2 142} _,746

b. Predictors: (Constant), My gender is:, My age is:, I experience Product D as:-Misleading:Not misleading, Product involvement, Issue involvement

c. Predictors: (Constant), My gender is:, My age is:, I experience Product D as:-Misleading:Not misleading, Product involvement, Issue involvement, moderator issue involvement D, moderator product involvement D

(34)

Based on the results discussed earlier in the model summary in table 17, the coefficients demonstrated in table 24 show no significant effect of age and gender. The effect of the three predictors is attributable to a strong positive effect of sensitivity to misleading product attributes, B=.85, p<.001. The positive coefficient indicates that respondents who are relatively sensitive to misleading product attributes, have higher purchase intentions regarding the stimuli products. Therefore, Hypothesis 1, “Deceptive product attributes have a negative effect on purchase intention”, is not confirmed.

When the interaction variables are added to the model, the results show that the misleading product attributes * product involvement moderator variable has a significant effect on purchase intention, B= -.30, p<.05. This means that Hypothesis 2, “The effect of deceptive product attributes is less negative when there is a higher level of consumer product involvement”, is not confirmed. Respondents with high product involvement demonstrate an almost linear purchase intention increase when sensitivity to misleading product attributes increases; high product involvement is associated with a strong positive effect of sensitivity to misleading product attributes.

In contrast, for respondents with low product involvement the purchase intention is especially high for respondents who are highly sensitive to misleading product attributes.

Hypothesis 3: “Consumer issue involvement moderates the effect of deceptive product attributes on purchase intention”. This hypothesis is not confirmed, because the effect of deceptive product attributes is less negative when there is a higher level of consumer issue involvement. There is no significant interaction effect of issue involvement * deceptive product attributes, B=.13, ns.

(35)

Table 24: Coefficients

, Model

Unstandardized

Coefficients Standardized Coefficients

t Sig. B Std. Error Beta 1 (Constant) 3,539 ,370 9,569 ,000 My age is: ,001 ,006 ,014 ,164 ,870 My gender is: -,322 ,218 -,122 -1,473 ,143 2 (Constant) ,507 ,553 ,916 ,361 My age is: ,004 ,005 ,049 ,720 ,473 My gender is: -,356 ,199 -,135 -1,786 ,076 Issue involvement ,112 ,125 ,093 ,891 ,375 Product involvement -,132 ,100 -,128 -1,328 ,186

Misleading product attributes ,848 ,096 ,578 8,795 ,000

3 (Constant) ,221 ,559 ,396 ,693

My age is: ,003 ,005 ,038 ,558 ,578

My gender is: -,338 ,197 -,128 -1,712 ,089

Issue involvement ,140 ,126 ,116 1,115 ,267

Product involvement -,135 ,100 -,130 -1,350 ,179

Misleading product attributes ,887 ,097 ,604 9,172 ,000

Moderator product involvement -,298 ,146 -,229 -2,037 ,043

Moderator issue involvement ,125 ,135 ,104 ,927 ,355

a. Dependent Variable: Purchase intention

The coefficient tables 25, 26, 27 and 28 on the next pages are based on tables 20, 21, 22 and 23 as discussed earlier and are also specified on each individual product. There is no significant effect of age and gender. The effect of the three predictors is attributable to a strong positive effect of awareness of misleading product attributes, p<.001. The positive coefficient indicates that

respondents who are aware of misleading product attributes (higher scores indicate less awareness), have lower purchase intentions regarding the stimuli products (higher scores indicate more purchase intention). When the interaction variables are added to the model, we see that the significant amount of extra explained variance of the moderators with Product C, cannot be attributed to either of the moderators. No significant moderating effects are found.

The role of product and issue involvement on the relationship between deceptive product attributes and the purchase intention of health food.